Mercurial > jhg

changeset 37:e45e75e22523

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Parse changesets from bundle's changelog group. Refactor Revlog to provide access to revision's raw content
author Artem Tikhomirov <tikhomirov.artem@gmail.com>
date Fri, 14 Jan 2011 00:49:58 +0100
parents 205f9b59b400
children 50dfc69c108e
files src/com/tmate/hgkit/console/Bundle.java src/com/tmate/hgkit/ll/HgBundle.java src/com/tmate/hgkit/ll/HgDataFile.java src/com/tmate/hgkit/ll/Revlog.java src/com/tmate/hgkit/ll/RevlogStream.java
diffstat 5 files changed, 63 insertions(+), 27 deletions(-) [+]
line wrap: on
line diff
--- a/src/com/tmate/hgkit/console/Bundle.java	Thu Jan 13 23:31:39 2011 +0100
+++ b/src/com/tmate/hgkit/console/Bundle.java	Fri Jan 14 00:49:58 2011 +0100
@@ -6,7 +6,9 @@
 import java.io.File;
 
 import com.tmate.hgkit.fs.DataAccessProvider;
+import com.tmate.hgkit.fs.RepositoryLookup;
 import com.tmate.hgkit.ll.HgBundle;
+import com.tmate.hgkit.ll.HgRepository;
 
 /**
  *
@@ -15,9 +17,17 @@
 public class Bundle {
 
 	public static void main(String[] args) throws Exception {
+		RepositoryLookup repoLookup = new RepositoryLookup();
+		RepositoryLookup.Options cmdLineOpts = RepositoryLookup.Options.parse(args);
+		HgRepository hgRepo = repoLookup.detect(cmdLineOpts);
+		if (hgRepo.isInvalid()) {
+			System.err.printf("Can't find repository in: %s\n", hgRepo.getLocation());
+			return;
+		}
 		File bundleFile = new File("/temp/hg/hg-bundle-a78c980749e3.tmp");
 		DataAccessProvider dap = new DataAccessProvider();
 		HgBundle hgBundle = new HgBundle(dap, bundleFile);
-		hgBundle.read();
+//		hgBundle.dump();
+		hgBundle.changes(hgRepo);
 	}
 }
--- a/src/com/tmate/hgkit/ll/HgBundle.java	Thu Jan 13 23:31:39 2011 +0100
+++ b/src/com/tmate/hgkit/ll/HgBundle.java	Fri Jan 14 00:49:58 2011 +0100
@@ -26,7 +26,29 @@
 		bundleFile = bundle;
 	}
 
-	public void read() throws IOException {
+	public void changes(HgRepository hgRepo) throws IOException {
+		DataAccess da = accessProvider.create(bundleFile);
+		try {
+			List<GroupElement> changelogGroup = readGroup(da);
+			byte[] baseRevContent = null;
+			for (GroupElement ge : changelogGroup) {
+				if (baseRevContent == null) {
+					// first parent is base revision, see bundlerepo.py
+					// if not prev: prev = p1 in bundlerevlog cons
+					baseRevContent = hgRepo.getChangelog().content(ge.firstParent());
+				}
+				int resultLen = 10000; // XXX calculate based on baseRevContent.length and ge.patches
+				byte[] csetContent = RevlogStream.apply(baseRevContent, resultLen, ge.patches);
+				Changeset cs = Changeset.parse(csetContent, 0, csetContent.length);
+				cs.dump();
+				baseRevContent = csetContent;
+			}
+		} finally {
+			da.done();
+		}
+	}
+
+	public void dump() throws IOException {
 		DataAccess da = accessProvider.create(bundleFile);
 		try {
 			LinkedList<String> names = new LinkedList<String>();
@@ -34,12 +56,12 @@
 				System.out.println("Changelog group");
 				List<GroupElement> changelogGroup = readGroup(da);
 				for (GroupElement ge : changelogGroup) {
-					System.out.printf("  %s %s %s %s; patches:%d\n", ge.node(), ge.firstParent(), ge.secondParent(), ge.cs(), ge.patches.size());
+					System.out.printf("  %s %s %s %s; patches:%d\n", ge.node(), ge.firstParent(), ge.secondParent(), ge.cset(), ge.patches.size());
 				}
 				System.out.println("Manifest group");
 				List<GroupElement> manifestGroup = readGroup(da);
 				for (GroupElement ge : manifestGroup) {
-					System.out.printf("  %s %s %s %s; patches:%d\n", ge.node(), ge.firstParent(), ge.secondParent(), ge.cs(), ge.patches.size());
+					System.out.printf("  %s %s %s %s; patches:%d\n", ge.node(), ge.firstParent(), ge.secondParent(), ge.cset(), ge.patches.size());
 				}
 				while (!da.isEmpty()) {
 					int fnameLen = da.readInt();
@@ -52,7 +74,7 @@
 					List<GroupElement> fileGroup = readGroup(da);
 					System.out.println(names.getLast());
 					for (GroupElement ge : fileGroup) {
-						System.out.printf("  %s %s %s %s; patches:%d\n", ge.node(), ge.firstParent(), ge.secondParent(), ge.cs(), ge.patches.size());
+						System.out.printf("  %s %s %s %s; patches:%d\n", ge.node(), ge.firstParent(), ge.secondParent(), ge.cset(), ge.patches.size());
 					}
 				}
 			}
@@ -103,9 +125,8 @@
 		public Nodeid secondParent() {
 			return Nodeid.fromBinary(header, 40);
 		}
-		public Nodeid cs() {
+		public Nodeid cset() { // cs seems to be changeset
 			return Nodeid.fromBinary(header, 60);
 		}
-
 	}
 }
--- a/src/com/tmate/hgkit/ll/HgDataFile.java	Thu Jan 13 23:31:39 2011 +0100
+++ b/src/com/tmate/hgkit/ll/HgDataFile.java	Fri Jan 14 00:49:58 2011 +0100
@@ -41,25 +41,6 @@
 		return content(TIP);
 	}
 
-	public byte[] content(Nodeid nodeid) {
-		int revision = content.findLocalRevisionNumber(nodeid);
-		return content(revision);
-	}
-	
-	/**
-	 * @param revision - repo-local index of this file change (not a changelog revision number!)
-	 */
-	public byte[] content(int revision) {
-		final byte[][] dataPtr = new byte[1][];
-		Revlog.Inspector insp = new Revlog.Inspector() {
-			public void next(int revisionNumber, int actualLen, int baseRevision, int linkRevision, int parent1Revision, int parent2Revision, byte[] nodeid, byte[] data) {
-				dataPtr[0] = data;
-			}
-		};
-		content.iterate(revision, revision, true, insp);
-		return dataPtr[0];
-	}
-
 	public void history(Changeset.Inspector inspector) {
 		if (!exists()) {
 			throw new IllegalStateException("Can't get history of invalid repository file node"); 
--- a/src/com/tmate/hgkit/ll/Revlog.java	Thu Jan 13 23:31:39 2011 +0100
+++ b/src/com/tmate/hgkit/ll/Revlog.java	Fri Jan 14 00:49:58 2011 +0100
@@ -34,7 +34,30 @@
 	public int getRevisionCount() {
 		return content.revisionCount();
 	}
+
+	/**
+	 * Access to revision data as is (decompressed, but otherwise unprocessed, i.e. not parsed for e.g. changeset or manifest entries) 
+	 * @param nodeid
+	 */
+	public byte[] content(Nodeid nodeid) {
+		int revision = content.findLocalRevisionNumber(nodeid);
+		return content(revision);
+	}
 	
+	/**
+	 * @param revision - repo-local index of this file change (not a changelog revision number!)
+	 */
+	public byte[] content(int revision) {
+		final byte[][] dataPtr = new byte[1][];
+		Revlog.Inspector insp = new Revlog.Inspector() {
+			public void next(int revisionNumber, int actualLen, int baseRevision, int linkRevision, int parent1Revision, int parent2Revision, byte[] nodeid, byte[] data) {
+				dataPtr[0] = data;
+			}
+		};
+		content.iterate(revision, revision, true, insp);
+		return dataPtr[0];
+	}
+
 	// FIXME byte[] data might be too expensive, for few usecases it may be better to have intermediate Access object (when we don't need full data 
 	// instantly - e.g. calculate hash, or comparing two revisions
 	// XXX seems that RevlogStream is better place for this class. 
--- a/src/com/tmate/hgkit/ll/RevlogStream.java	Thu Jan 13 23:31:39 2011 +0100
+++ b/src/com/tmate/hgkit/ll/RevlogStream.java	Fri Jan 14 00:49:58 2011 +0100
@@ -284,7 +284,8 @@
 
 	// mpatch.c : apply()
 	// FIXME need to implement patch merge (fold, combine, gather and discard from aforementioned mpatch.[c|py]), also see Revlog and Mercurial PDF
-	private static byte[] apply(byte[] baseRevisionContent, int outcomeLen, List<PatchRecord> patch) {
+	// FIXME why 2 arrays (rv and tempBuf)???. Think over in-place merging (perhaps some sparse byte chunk list?) - to minimize mem use.
+	/*package-local for HgBundle; until moved to better place*/static byte[] apply(byte[] baseRevisionContent, int outcomeLen, List<PatchRecord> patch) {
 		byte[] tempBuf = new byte[outcomeLen]; // XXX
 		int last = 0, destIndex = 0;
 		for (PatchRecord pr : patch) {